Photochemistry and fluorescence of bacteriorhodopsin excited in its

Marr, A. M. S., Donald, A. S. R., Watkins, W. M., &. Morgan, W. T. J. (1967) Nature (London) 215,1345-1349. Mourant, A. E. (1946) Nature (London) 158,...
0 downloads 0 Views 672KB Size
Biochemistry 1981, 20, 205-209 Heyderman, E., & Monaghan, P. (1979) Invest. Cell Pathol. 2, 119-122. Iseki, S., Masaki, S . , & Shibasaki, K. (1957) Proc. Jpn. Acad. 33, 492-497. Judd, W. J., Steiner, E. A,, Friedman, B. A., & Oberman, H. A. (1978) Transfusion 18, 436-440. Kent, S. P. (1964) J . Histochem. Cytochem. 12, 591-599. Kobata, A. (1975) Kagaku (Kyoto) 30, 192-198. Lemieux, R. U. (1978) Chem. SOC.Rev. 7,423-452. Lemieux, R. U., Bundle, D. R., & Baker, D. A. (1975) J. Am. Chem. SOC.97,4076-4083. Lemieux, R. U., Baker, D. A., & Bundle, D. R. (1977) Can. J . Biochem. 55, 507-5 12. Lemieux, R. U., Bundle, D. R., & Baker, D. A. (1979) U S . Patent 4 137401. Lemieux, R. U., Bock, K., Delbaere, L. T. J., Koto, S., & Rao, V. S. (1980a) Can. J . Chem. 58,631-653. Lemieux, R. U., Le Pendu, J., & Hindsgaul, 0. (1980b) J . Antibiot. 32, S21-31. Lloyd, K. O., Kabat, E. A., & Licerio, E. (1968) Biochemistry 7, 2976-2990. Marcus, D. M., & Grollman, A. P. (1966) J . Immunol. 97, 867-87 5. Marr, A. M. S., Donald, A. S . R., Watkins, W. M., & Morgan, W. T. J. (1967) Nature (London) 215,1345-1349. Mourant, A. E. (1946) Nature (London) 158, 237-238. Perera, D. R., Weinstein, W. M., & Rubin, C. E. (1975) Hum. Pathol. 6, 157-215. Potapov, M. I. (1970) Probl. Haematol. 11, 45. Potapov, M. I. (1976) Vox Sang. 30, 211-213. Prohaska, R., Schenkel-Brunner, H., & Tuppy, H (1978) Eur. J . Biochem. 84, 161-166.

205

Race, R. R., & Sanger, R. (1972) Haematologia 6, 63-71. Rege, V. P., Painter, T. J., Watkins, W. M., & Morgan, W. T. J. (1964) Nature (London) 204, 740-742. Rovis, L., Anderson, B., Kabat, E. A,, Greuzo, F., & Liao, J. (1973) Biochemistry 12, 5340-5354. Schenkel-Brunner, H., & Prohaska, R. (1978) Proc. Int. Congr. Int. SOC.Hematol., 17th-Int. SOC.Blood Transfusion, 15th, Abstracts p 646. Schiffman, G., Kabat, E. A., & Thompson, W. (1964) Biochemistry 3, 113-120. Schwarting, G. K., & Marcus, D. M. (1979) Clin. Immunol. Immunopathol. 14, 121-129. Shen, L., Grollman, E. F., & Ginsburg, V. (1968) Proc. Natl. Acad. Sci. U.S.A. 59, 224-230. Smith, E. L., McKibbin, J. M., Karlsson, K.-A,, Pascher, I., & Samuelson, B. E. (1975) J . Biol. Chem. 250,6059-6064. Stellner, K., Watanabe, K., & Hakomori, S. (1973) Biochemistry 12, 656-661. Szulman, A. E. (1962) J . Exp. Med. 115, 977-996. Szulman, A. E., & Marcus, D. M. (1973) Lab. Invest. 28, 565-574. Ueyama, R. (1939) Hanzaigaku Zasshi 13, 51-64. Watkins, W. M., & Morgan, W. T. J. (1957) Nature (London) 180, 1038-1040. Weinstein, W. M., & Lechago, J. (1977) Gastroenterology 73, 765-767. Wilbrant, R., Tung, K. S. K., Deodhar, S . D., Nakamoto, S . , & Kolff, W. J. (1969) Am. J . Clin. Pathol. 51, 15-23. Zopf, D. A., & Ginsburg, V. (1975) Arch. Biochem. Biophys. 167, 345-350.

Photochemistry and Fluorescence of Bacteriorhodopsin Excited in Its 280-nm Absorption Bandt 0. Kalisky, J. Feitelson, and M. Ottolenghi*

ABSTRACT:

Photochemical and fluorescence studies are carried out, exciting bacteriorhodopsin (BR) in its 280-nm absorption band. The data indicate that energy transfer takes place, with a quantum yield of 0.7-0.8, from excited tyrosines and tryptophans to the retinyl chromophore. All of the tyrosine and five to six tryptophan residues are completely quenched by the transfer process while one tryptophan is unquenched and one is partially (-80%) quenched. Energy transfer to the chromophore leads to a photocycle identical with that triggered in (light adapted) bacteriorhodopsin by excitation within the

visible absorption bands of the chromophore. The emissive properties of BR in the intact membrane are found equal to those of a Triton X-100 solubilized BR monomer. The energy transfer data are discussed in terms of the available amino acid sequence and the electron density map of bacteriorhodopsin. Although such data cannot suggest a single fit between the sequence and the density map (one out of the 7! = 5040 possibilities), they do provide a criterion for testing any specific model for the structure of bacteriorhodopsin.

Bacteriorhodopsin, the protein pigment of the purple membrane of Halobacterium halobium, is responsible for a light-driven proton pump which leads to ATP synthesis. [For a review, see Stoeckenius et al. (1978).] Analogous to visual pigments [for a review, see Ottolenghi (1980)], the spectrum of bacteriorhodopsin in the visible and in the near-UV is

characterized by three absorption bands (Becher et al., 1978). The a band (with an extinction of E = 63 000 M-' cm-I) peaks, in light-adapted (all trans) bacteriorhodopsin (BRf570),at 570 nm. A lower extinction band ( E = 13 000 M-' cm-'), lacking a clear maximum, covers the range between 450 and 300 nm (/3 band). In the UV spectrum, an intense ( E i= 7 5 000 M-' cm-') band is observed, peaking at 280 nm (y band). Both CY and /3 bands are exclusively due to the retinyl chromophore, which is attached to the protein via a protonated Schiff-base linkage. In keeping with the characteristic protein fluorescence induced by excitation in the near-UV spectrum (Bogomolni

7 From the Department of Physical Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel. Received May 19, 1980. This work was supported by grants from the Israeli Commission for Basic Research and from the US.-Israel Binational Science Foundation.

0006-296018 110420-0205$01 .OO/O 0 1981 American Chemical Society

206

KALISKY, FEITELSON, A N D OTTOLENGHI

BIOCHEMISTRY

et al., 1978), the y band is attributed to the superimposed absorptions of the retinyl moiety and of the aromatic protein residues, tyrosine and tryptophan. Becher et al. (1978) have recently shown that the latter are the major contributors to the absorption at 280 nm, with only 10% of the total extinction being due to the retinyl polyene. Extensive investigations have been carried out on the photochemistry of BR'570 induced by excitation within the a and p bands [see Stoeckenius et al. (1979) and Ottolenghi (1980) for recent reviews]. In the present work we have studied the photochemical and emissive properties of BR'J70 following excitation within the 280-nm (y) band. The results indicate that substantial energy transfer takes place from excited aromatic residues to the polyene moiety, triggering the characteristic photocycle of BR'570. An analysis of the fluorescence data bears on the tertiary structure of the bacteriorhodopsin molecule.

-

Experimental Procedures Experiments were carried out at room temperature with aqueous purple-membrane suspensions of M I Halobacterium halobium prepared according to previously described procedures (Oesterhelt & Stoeckenius, 1974). Work with bacteriorhodopsin solubilized with Triton X-100 at pH 5 (0.1 M acetate buffer) followed the procedures described by Stoeckenius and co-workers (Reynolds & Stoeckenius, 1977; R. Casadio, H. Gutowitz, P. Mowery, M. Taylor, and W. Stoeckenius, unpublished results). At pH 7, we employed the procedure of Becher & Ebrey (1976) using 0.025 M phosphate buffer. Fluorescence measurements in membrane suspensions were performed in the presence of 40% sucrose to minimize light scattering artifacts (Sherman & Caplan, 1975). Solutions of naphthalene, a Fluka (Puriss) product, were prepared in spectrograde cyclohexane. NATA (N-acetyltryptophanamide), obtained from Miles-Yeda (Reovoth), and glycyltyrosine, an ICN-National Biochemicals Division product, were used without further purification. A Laser-Associates Q-switched, frequency quadrupled, Nd laser was employed for pulsed (20 ns) excitation at 265 nm in a previously described experimental setup (Goldschmidt et al., 1976). A conventional fluorometer, including a Spex Minimate double monochromator for excitation and a f = 4, 500 nm, focal length Bausch & Lomb monochromator for detection at 90°, was employed for continuous excitation experiments. The emission, detected by a 5256s EM1 photomultiplier, was analyzed by a Brookdeal-Ortec photon-counting system (Lasser & Feitelson, 1973). The apparatus for determination of the fluorescence lifetimes has been recently described (Barboy & Feitelson, 1978). Briefly, the lifetimes in 2.5 X 10" M BR solutions were measured by illumination (at 265 or 292 nm) with a TRW nanosecond light source. The lamp response and the fluorescence were monitored by a DUVP Amperex photomultiplier coupled to a sampling oscilloscope and, hence, to a computer of average transients (CAT). For minimization of errors due to electronic drifts, the signals from the sample and from the scattering solutions were alternately measured for 10-s intervals and the data stored in separate compartments of the CAT. The lifetimes were obtained by convoluting the lamp response curve with a mono- or double-exponential decay function and fitting it to the experimental decay curve. Results and Discussion Pulsed Laser Photolysis at 265 nm. Photochemical experiments exciting BR'570 within its UV absorption band were carried out by using the 265-nm pulse of the frequency-

350

400

450

500

550

600

650

X,nm

FIGURE 1: Characteristic oscillogram and difference spectrum showing the generation of the M410phototransient of BIJ570following excitation with the 265-nmlaser pulse (unbuffered purple membrane suspensions, pH 5.5, at 20 "C).

quadrupled Nd laser. Excitation leads to transient phenomena identical with those previously reported following irradiation within the main (a)or secondary (@)absorption bands [for a review, see Ottolenghi (1980)l. For example, the decay at 420 nm and the corresponding difference spectrum obtained following 265-nm excitation (see Figure 1) are both characteristic of the familiar spectral and kinetic decay patterns associated with the M412phototransient (Goldschmidt et al., 1976). It is therefore suggested that UV irradiation triggers photochemical processes identical with those induced by visible light. A similar conclusion was previously reached for visual (bovine) rhodopsin (Rosenfeld & Ottolenghi, 1977). The cycling quantum yield for 265-nm excitation (4;:) was determined by using the naphthalene triplet state in cyclohexane as the reference. Use was made of the expression (Goldschmidt et al., 1977) :4; = ~R(A~R/A~BR)(ADBR)/ (ADT) (1 - lO%)/l- 10%~)where & = 0.71 is the reference (naphthalene) triplet yield (Amand & Bensasson, 1975) and AcR and AQR represent the differences in extinction coefficients between the molecules photolyzed (naphthalene and BR'570) and their respective photoproducts (the triplet and M410). DR and DBR represent the initial absorbance of the reference and of the BR'570 solution at 265 nm while ADT and AD, are the photoinduced changes in absorbance due to the generation of the triplet and the M410 intermediate, respectively. Experiments were carried out for two BR'570 concentrations (DBR = 0.45 and DBR = 0.81) by using low intensity pulses so as to assure linearity between AD and the pulse intensity was measured at 415 nm and (Goldschmidt et al., 1977). ADBR at 600, 580, and 565 nm. The results of a set of ten runs, varying the detection wavelength and the pulse intensity, yielded a consistent value of 4:; = 0.19 f 0.02. The absorbance of BR'570 at 265 nm is due to the superimposed contributions, and t:?, of the retinyl chromophore and of the opsin aromatic amino acids, respectively. The BR'570photocycle may therefore be induced both by direct light absorption by the retinyl polyene and by energy transfer to the latter from the excited amino acids. Denoting the cycling quantum yield by y, we may write

e

(1) &2 = f +~ ~ ( -1 J ~ E T + is the fraction of the 265-nm

where f = t ~ f ' / / " ~ ' ' quanta absorbed by the polyene and 4ETis the yield of energy transfer to the latter from the excited aromatic residues. The application of the above expression is associated with the exact determination off. On the basis of the spectrum

BACTERIORHODOPSIN PHOTOCHEMISTRY AND FLUORESCENCE

of a free protonated Schiff base in solution (PRBS), Becher et al. (1978) have recently shown that the contribution of the retinyl moiety to the total BR'57O extinction at 280 nm: ;e( = 75 000 M-' cm-') amounts to 6:;' = 8000 M-' cm-'. Since for PRSB the absorbance at 265 nm is essentially identical with that at 280 nm, the procedure of Becher et al. (1978) yields a trt value close to 8000 M-' cm-' (f = 0.12) also at 265 nm. Using :4; = 0.19 and y = 0.25 (see below), we obtain @ET = 0.727. A similar result ($JET= 0.730) is obtained with the value5::e = 7045 M-' cm-' cf = 0.106), calculated by using the spectrum of the opsin recorded after photobleaching the purple membrane in the presence of hydroxylamine, followed by extraction of the produced oxime (Becher et al., 1978). A third approach involves the use of a NATA/Tyr-Gly solution identical in its tryptophan-tyrosine composition with that of bacteriorhodopsin [ 11 tyrosines and 7 tryptophans according to Ovchinikov et al. (1979) or 11 tyrosines and 8 tryptophans according to Khorana et al. (1979)l. At 280 nm, such solutions are characterized (Mihalyi, 1968) by e = 55 OOO M-' cm-I as compared with 6;: = 75000 M-' crn-'. The difference between the two values was attributed by Becher et al. (1978) to environmental effects on the absorption of the aromatic amino acids in the opsin. Assuming that the relative contributions of such effects are identical at 280 and 265 nm, we have estimated' : :e by using the respective extinction in the NATA/Tyr-Gly solution and the above value of. ' : :e Although the result,'::e = 15 700 M-' cm-l (f = 0.22-0.23), is substantially different from the values yielded by the previous procedures, the calculated energy transfer quantum yield ($ET = 0.69) is close to those calculated above. While the first of the three procedures assumes that the relative absorptions at 265 and 280 nm are identical for a free PRSB and for the retinyl chromophore in the opsin, the second overlooks structural changes in the opsin which might be induced by bleaching and the third neglects any wavelength dependence of the environmental effects in the opsin. Nevertheless, in view of the reasonable agreement between the values of yielded by the three methods, we consider $ET = 0.71 f 0.02 as a good estimate for the yield of energy transfer to the chromophore. We should finally mention that in the above calculations of @ET we have used the value y = 0.25 which was measured for visible (530 nm) excitation, populating the lowest excited state (S,) of the chromophore (Goldschmidt et al., 1977). This approximation involves two assumptions: first, as it is almost always the case, the excitation of the chromophore to higher excited states (by direct adsorption at 265 nm or by energy transfer from the protein) excited results in the quantitative population of the lowest, SI, state, and, second, the quantum yield in SI is independent of the particular vibronic levels initially populated. This assumption is justified by the independency of y on the excitation wavelength within the main 570-nm absorption band, a phenomenon interpreted in terms of the quantitative population of a common minimum along the excited-state potential surface connecting BR'570 and its primary photoproduct K610 (Rosenfeld et al., 1977; Hurley et al., 1977). Fluorescence Measurements. The emission of BR'570, excited in the UV band, was studied by using continuous and pulsed excitation to gain further insight into the details of the = 320 nm) energy transfer process. Emission spectra (A, obtained by continuous irradiation at 265 and 292 nm agreed with that reported by Bogomolni et al. (1978) for 280-nm excitation. There are several points indicating that only light absorbed by tryptophan residues is responsible for the observed emission. First, identical emission spectra are obtained for

VOL. 20, NO. 1, 1981

207

excitation at 292 nm, where only tryphotopan absorbs, and for 265-nmexcitation where there is a considerable tyrosine contribution to the absorption. Second, the relative emission intensities due to 265- and 292-nm excitation in a dilute NATA solution are the same as in a comparable BR1570solution. This not only indicates that the emission is solely due to tryptophan residues but also that the excitation of tyrosines in BR'570 does not lead to a tryptophan fluorescence via energy transfer. An additional argument favoring the same conclusions may be made on the basis of the measured emission decay kinetics in BR'570 which were the same, within experimental error, for 265- and 292-nm excitation. Thus, the deconvolution analysis yielded in both cases a double-exponential decay with half-lives of T , = 3.5-5.0 ns and T~ = 0.5-0.9 ns. For the determination of the fluorescence quantum yield, we used the values for the extinction coefficients of the aromatic residues and of the nonemissive retinyl chromophore, estimated as discussed above. In calculating the value for 265-nmexcitation, we assumed, as previously concluded, that light absorption by tyrosine residues does not lead to emission. The value obtained, 4f = 0.023 f 0.002, was essentially the same for both excitation wavelengths, in agreement with the assumption that only tryptophan residues emit. For N tryptophan residues present in bacteriorhodopsin, we have c#q = ( C n & ) / N , where n, is the number of molecules characterized by the same emission lifetime, T i = I # J ~ TT~ ~ , is the radiative lifetime, and 4, is the emission quantum yield of a tryptophan residue belonging to the ith group. The previously reported lifetime data indicate that the emitting molecules ( T ~> 0.1 ns) can be divided into two major groups with T~ = 3.5-5.0 ns and T~ = 0.5-0.9 ns, respectively. In applying the above equation, we first substituted N = 7 (Ovchinikov et al., 1979) and df = 0.023. For T~ we have used ~ 22.3 ns) characteristic of aqueous tryptophan the value ( T = (NATA) solutions (Chen et al., 1969; Grinvald et al., 1975). We are left with the two unknown parameters, nl and n2. Since it can readily be verified that any value of nl which is greater than 1 will lead to the violation of the elementary relation nl& I4fN,we conclude that nl = 1. This conclusion is consistent only with n2 = 1, implying that only one tryptophan is responsible for the long lifetime emission and a second one for the subnanosecond component, while the remaining five residues are undetectable by the present experiments. This conclusion is independent on the specific choice of and of T~ in the 3.5-5.0-11s and 0.5-0.9-11s ranges, respectively. An identical picture is obtained by using N = 8 (Khorana et al., 1979) except that, in this case, the possibility of n2 = 2 cannot be absolutely ruled out. At 265 nm the contribution to the total protein absorbance by the 7 tryptophan and the 11 tyrosine residues can be approximated from the data of Mihalyi (1968) in water as 0.77 and 0.23, respectively. With 11 fully quenched tyrosines, five (Ovchinikov et al., 1979) or six (Khorana et al., 1979) fully quenched tryptophans, one partially quenched (-80%) tryptophan, and one unquenched tryptophan, we obtain 0.86-0.88 as the yield of all quenching pathways of the excited aromatic amino acids in bacteriorhodopsin. This value constitutes an = 0.71 f 0.02, upper limit to &T. In view of the value obtained above from the photochemical data, it is thus implied that about 80% of the quenching is due to energy transfer to the chromophore. Continuous excitation fluorescence experiments were also carried out in dark-adapted BR solutions solubilized by Triton X-100. Solubilization was carried out at either pH 5 or pH 7 (see Experimental Procedures). We used a 1:32 (w/w)

208

BIOCHEMISTRY

protein to detergent ratio and performed the experiments 50 h after adding the detergent so as to allow almost complete solubilization to take place (Reynolds & Stoeckenius, 1977; Dencher & Heyn, 1978a,b). Since Triton X-100 absorbs a substantial fraction of the exciting light and is also highly fluorescent when illuminated near 280 nm, we have used excitation at 300 nm where the detergent absorption is relatively low. The emission intensity of a BR-solubilized solution at 340 nm was compared to that containing only the detergent and to an aqueous sucrose solution of purple membranes. After correcting for the large detergent background emission (-90%) and evaluating the exact absorbance at 300 nm (correcting for scattering artifacts), we obtained the values 0.8 f 0.2 and 1.1 f 0.2 for the ratio between the BR emission intensity in the detergent micelle and in the intact membrane fragments at pH 5 and 7, respectively. Experiments carried out in a NATA solution in the prescence of the same amount of Triton X-100 indicated a small (-25%) increase in the emission in the presence of the detergent, implying that no quenching of tryptophan fluorescence by Triton X- 100 takes place. It should be pointed out that we preferred the use of Triton X-100 in spite of its strong fluorescence to that of the nonfluorescent octyl glucoside. Solubilization with the latter has been reported to induce an increase in the BR fluorescence which is not restored upon reconstitution of the membrane (Dencher & Heyn, 1978a,b). It is likely that this effect is due to an irreversible damage to the bacteriorhodopsin (e.g., bleaching). Structural Implications. It is worthwhile examining the above conclusions in terms of the available information on the structure of bacteriorhodopsin in the purple membrane. On the basis of electron microscopy data, Henderson & Unwin (1975) have shown that each BR molecule is composed of seven a-helical segments spanning the membrane. The molecules are arranged in the form of trimers in a two-dimensional hexagonal lattice (Figure 2). The amino acid sequence in bacteriorhodopsin has recently been obtained by Ovchinikov et al. (1979) and by Khorana et al. (1979) who presented a detailed model describing the composition of each a-helical segment, including the location of the retinyl chromophore. An attempt has recently been made to correlate the a-helical segments in the electron density map with those suggested on the basis of the primary structural analysis (Engelman et al., 1980). For an analysis of our energy-transfer data in terms of the above structural information, we apply Forster's mechanism (Forster, 1966), which leads to the expression q5kT = R,"/(R," Rod)for the yield of energy transfer from the ith aromatic residue to the chromophore. R, is the donoracceptor distance, the value of R for which q5bT = 0.5, is given by Ro (A) and Ro, = (JK2q5pn-4)*/6(9.79 X lo3), The parameter J (cm3 M-l), the overlap function between the emission of the donor and the ( p band) absorption of the acceptor, is given by J = SF(A)t(A)X4 dX/IF(A)dA where F(X) represents the wavelength-dependent fluorescence intensity and E( A) represents the extinction coefficient of the acceptor. For all residues we assumed a value of n = 1.4 for the refractive index (Cheng= 2 / 3 for the square of the Wen & Stryer, 1972) and of orientation factor K . The latter assumption, which is strictly valid only when the donor and acceptor rotation rates are fast compared to those of energy transfer, is considered a fair approximation in proteins (Cheng-Wen & Stryer, 1972; Gafni, 1978). In all cases we assume that @, the fluorescence yield of each tryptophan residue in the absence of the chromophore acceptor, is identical with that of the long-lived emitting

+

KALISKY, FEITELSON, A N D OTTOLENGHI

FIGURE 2: (Top) Schematic representation of the seven a-helical segments of bacteriorhodopsin in the purple membrane according to Ovchinikov et al. (1979). The circles and squares represent tyrosine and tryptophan residues, respectively, numbered according to the sequence of Khorana et al. (1979). R represents the retinyl moiety attached to lysine-41. (Bottom) Schematic representation of the electron density map of bacteriorhodopsin according to Henderson & Unwin (1975). The circles represent the helical segments, and the specific fit of the amino acid sequence with the density map is that suggested as mast probable by Engelman et al. (1980). The orientation of the retinyl moiety is suggested by the present energy transfer data.

tryptophan ( T ~= 3.5-5.0 ns). This assumption is in keeping with the models of Ovchinikov et al. (1979) and of Khorana et al. (1979) which place all tryptophans in the interior of the membrane (see Figure 2). By using the above approximations, the value of Ro for transfer from tryptophans to the chromophore is calculated as Ro = 25 A. The complete set of Ri for all aromatic residues is determined by the overall protein structure. In other words, the values of Rimay be derived from the exact correlation (one out of 7! = 5040 possibilites) between the a-helical segments in the electron density diagram (see Figure 2) and those described in the molecular model (Ovchinikov et al., 1979; Khorana et al., 1979). Although an examination of such correlations in terms of our energy-transfer data shows that most structures are inconsistent with the experimental results, it is impossible to point out a specific configuration on the basis of the fluorescence data alone. It is evident, however, that any specific model proposed must be compatible with the main conclusions of the present work, namely, (a) the complete quenching of all tyrosine residues, (b) the complete quenching of all tryptophan residues, with the exception of one un-

BACTERIORHODOPSIN PHOTOCHEMISTRY A N D FLUORESCENCE

quenched and one (or two) partially (-80%) quenched, and (c) the identity, within our experimental limits, between the quenching (energy transfer) phenomena in the intact lattice and those observed after treatment with Triton X-100. The latter process involves the formation of a monomeric BR micelle, free of intermolecular interactions (Becher & Ebrey, 1976; Reynolds & Stoeckenius, 1977). It is thus implied that all energy-transfer phenomena from tyrosines and tryptophans are solely accounted for by the chromophore from the same BR molecule. Figure 2 shows a specific model recently selected as most probable by Engelman et al. (1980). The suggestion was based on criteria of connectively of the nonhelical link regions, charge neutralization, and total scattering density per helix. It can be verified that the model meets the three requirements mentioned above if the chromophore, attached to lysine 41 on helix B, is placed close to helix G as shown in Figure 2. Further structural restrictions are imposed by the fact that the complete quenching of five to six tryptophan residues implies that their respective distances from the chromophore cannot exceed -20 A, a value which corresponds to -90% energy transfer. As shown in Figure 2, the model of Ovchinikov et al. (1979) places seven tryptophan residues in a plane which is 10 A from the external surface of the membrane. Since the total membrane width is -40 A, it is thus implied that the chromophore plane lies at least 10 A from the cytoplasmic surface. Within the limits imposed by the approximation k? = 2/3, adoption of the model of Engelman et al. (1980) (Figure 2) suggests that the unquenched tryptophan residue lies on helix E (Trp-137 or -138) which is the most distant from the chromophore. The partially (-80%) quenched residue may be the second tryptophan of helix E (with an unsuitable orientation) or one of those on helix F (Trp-189). Total quenching is predicted for at least five tryptophan residues and also for all tyrosine residues, in agreement with the experimental data. Quenching of the tyrosines is accounted for by energy transfer to neighboring tryptophans, a process characterized by Ro= 17 A (Berlman, 1973). Thus, tryptophans will act as efficient mediators for the transfer of tyrosine excitation energy to the chromophore, even from residues which are too distant for direct transfer to the retinyl moiety. An examination of Figure 2 shows that the model does not imply any solubilization effects on the energy transfer phenomena.

-

References Amand, B., & Bensasson, R. (1975) Chem. Phys. Lett. 34, 44-48. Barboy, N., & Feitelson, J. (1978) Biochemistry 17, 4923-4926. Becher, B., & Ebrey, T. G. (1976) Biochem. Biophys. Res. Commuh. 69, 1-6. Becher, B., Tokunaga, F., & Ebrey, T. G. (1978) Biochemistry 17,2293-2300.

VOL. 20, N O . 1 , 1981

209

Berlman, I. B. (1973) Energy Transfer Parameters of Aromatic Compounds, Academic Press, New York. Bogomolni, R. A., Stubbs, L., & Lanyi, J. K. (1978) Biochemistry 17, 1037-1041. Chen, R. F., Edelhoch, H., & Steiner, R. F. (1969) in Physical Principles and Techniques of Protein, Part A (Leach, S. J., Ed.) pp 171-244, Academic Press, New York, NY. Cheng-Wen, W., & Stryer, L. (1972) Proc. Natl. Acad. Sci. U.S.A. 69, 1104-1 108. Dencher, N. A., & Heyn, P. M. (1978a) in Energetics and Structure of Halophilic Microorganisms (Caplan, S . R., & Ginzburg, M., Eds.) pp 233-238, Elsevier/North-Holland Biochemical Press. Dencher, N. A., & Heyn, P. M. (1978b) FEBS Lett. 96, 3 22-3 26. Engelman, D. M., Henderson, R., McLachlan, A. D., & Wallace, B. A. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 2023. Forster, Th. (1966) in Modern Quantum Chemistry, Istanbul Lectures (Sinnanoglu, O., Ed.) Section 111-B, pp 93-137, Academic Press, New York. Gafni, A. (1978) Annu. Rep. Prog. Chem., Sect. A 7 5 , 5 2 4 . Goldschmidt, C. R., Ottolenghi, M., & Korenstein, R. (1976) Biophys. J. 16, 838-843. Goldschmidt, C. R., Kalisky, O., Rosenfeld, T., & Ottolenghi, M. (1977) Biophys. J . 17, 179-183. Grinvald, A., Schlessinger, J., Pecht, I., & Steinberg, I. Z. (1975) Biochemistry 14, 1921-1928. Henderson, R., & Unwin, P. N. T. (1975) Nature (London) 257,28-32. Hurley, J. B., Ebrey, T. G., Honig, B., & Ottolenghi, M. (1977) Nature (London) 270, 540-542. Khorana, H. G., Gerber, G. E., Herlihy, W. C., Gray, C. P., Anderegg, R. J., Nihel, K., & Biemann, K. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 5046-5050. Lasser, N., & Feitelson, J. (1973) J . Phys. Chem. 77, 1011-1016. Mihalyi, E. (1968) J . Chem. Eng. Data 13, 179-182. Oesterhelt, D., & Stoechenius, W. (1974) Methods Enzymol. 31, 667-678. Ottolenghi, M. (1980) Adu. Photochem. 12, 97-200. Ovchinikov, A. Yu., Abdulaev, N. G., Feigina, M. Yu., Kiselev, A. V., & Labanov, N. A. (1979) FEBS Lett. 100, 219-224. Reynolds, J. A., & Stoeckenius, W. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 2803-2804. Rosenfeld, T., & Ottolenghi, M. (1977) in Research in Photobiology (Castellani, A., Ed.) pp 667-675, Plenum Press, New York. Rosenfeld, T., Honig, B., Ottolenghi, M., Hurley, J., & Ebrey, T. G. (1977) Pure Appl. Chem. 49, 341-351. Sherman, W. V., & Caplan, S. R. (1975) Nature (London) 258, 766-768. Stoeckenius, W., Lozier, R. H., & Bogomolni, R. A. (1979) Biochim. Biophys. Acta 505, 215-278.